Artificial Intelligence And Judicial Decision- Making - An Examination Of India's Legal Framework
Author: LawSphereX
Category: Technology, Cyber & Privacy
Published:
Updated:
ARTIFICIAL INTELLIGENCE AND JUDICIAL DECISION-
MAKING: AN EXAMINATION OF INDIA'S LEGAL
FRAMEWORK
Balaji Babu D, Rajiv Gandhi School of Intellectual Property Law, IIT Kharagpur
ABSTRACT
This article explores the impact of engagement of Artificial Intelligence in the judicial system in India and its ramifications for constitutional governance. While the existing AI tools including SUPACE and SUVAS have exhibited tremendous impact in efficiency and case management, extending their role raises serious concerns regarding transparency, accountability and fairness in judicial decision-making. The article undertakes to study the risks associated with algorithmic opacity and the possibility of bias being perpetuated through dependence on historical judicial data, which may adversely affect the guarantee of equality before law under Article 14. It further studies the consequences of such technological integration on the right to fair procedure under Article 21, particularly in light of principle of reasoned decisions.
Keywords: Artificial Intelligence, Judicial Decision Making, Algorithmic Bias, Judicial Accountability, Due Process.
I. INTRODUCTION
The continuing challenge to the Indian judiciary is the pending cases in millions before various courts across the country and the delayed justice leading to denial of justice for ordinary citizens has made the technological solutions appealing. 1 The recent initiative by the supreme court to explore AI applications such as Supreme Court Portal for Assistance in Court Efficiency (SUPACE) and Supreme Court Vidhik Anuvaad Software (SUVAS ) offers speedy disposal and wider accessibility to legal information2. Yet, perturbing questions exist beneath this positive vision. Who ensures fair trial when the judicial decisions are influenced by algorithms? How can a computer system comprehend the intricate circumstances of legal disputes between humans? What will be the consequence of perpetuation of discrimination by AI, owing to judicial data being discriminative?
This paper undertakes to study whether AI can help reduce the burden of Indian courts or whether it jeopardizes the constitutional rights of fairness, equality and human dignity.
II. STATEMENT OF THE PROBLEM
The judicial system is facing multiple challenges: a case pendency of 5 crore, lack of judges owing to delay in appointment, and availability of access to justice only to the wealthy individuals who can afford expensive legal representation. 3 While the technology promises solutions, there exist risks in employing AI in the justice delivery system. The judges and litigants cannot comprehend as to why an algorithm gave a specific decision resulting in AI systems operating as “black boxes”. This opaqueness violates the legal principle of reasoned decision and impedes the opportunity to the litigant to challenge the decisions. 4 AI absorbs the biases within the past decisions when it is trained on historical judicial data.
The algorithm will likely emulate the discriminative decisions against certain communities, causing violation of equality before law as espoused within Article 14 of the Constitution of India. 5 The 1 Tanmay Nirmal, Artificial Intelligence, E-Justice, and Sustaining the Rule of Law—Transforming Indian Justice for the Digital Era (Nat’l e-Governance Div. Dec. 2025), https://negd. gov. in/wp-content/uploads/2025/12/Readyto-upload-case-study-AI-E-Justice-Rule-of-Law-1-1. pdf. 2 Nivash Jeevanandam, Artificial Intelligence in Judicial Processes: Transforming India’s Legal System, INDIAAI (Feb. 7, 2025), https://indiaai. gov. in/article/ai-in-judicial-processes-transforming-india-s-legal-system.
3 Id.
4 Victoria Hendrickx, The Judicial Duty to State Reasons in the Age of Automation? The Impact of Generative AI Systems on the Legitimacy of Judicial Decision-Making, 44 ERASMUS L.REV. (2024)
5 Sahil Girhepuje et al., Are Models Trained on Indian Legal Data Fair?, arXiv (Mar. 13, 2023),
https://arxiv.org/abs/2303.07247.
western legal data has been predominantly used in development of AI legal technology. Application of such system to unique social conditions and distinct constitutional framework of India, may result in inaccurate decisions unsuitable for Indian conditions.6 The principle of judicial independence is eroded when the reliance is placed on algorithmic recommendations. Judges under disposal targets may defer to AI usage rather than relying on self judgement emanating from wisdom and legal principles.7
III. RESEARCH FRAMEWORK
Research Questions
1. What are the AI tools utilized by Indian courts and their legally permissible role in arriving
at judicial decisions?
2. How does AI adoption in justice delivery system affect fundamental constitutional
guarantees of the right to equality and due process under Articles 14 and 21? Research Objectives
1. To identify the existing AI tools in Indian judiciary and examine their operation against
judicial accountability and transparency.
2. To analyse the consequences of adopting AI assistance in judicial decision making on right
to equality and due process under Articles 14 and 21.
IV. LITERATURE REVIEW
The present research aims to address the distinct gaps revealed by the existing scholarship on AI adoption in Indian judiciary. The study by Bhaswati Talukdar, examines in detail about the AI tools like SUPACE and AI committee of supreme court, with focus on efficiency gains, judicial discretion and algorithmic bias. Though the article is pertinent to Indian context and provides elaborated information on adoption of AI, it overlooks the implications on judicial 6 Kushagra Tiwari, AI Training Opt-Outs Reinforce Global Power Asymmetries, INDIAN J. L. & TECH. BLOG (2024), https://forum.nls.ac.in/ijlt-blog-post/ai-training-opt-outs-reinforce-global-power-asymmetries/ (last visited Apr. 12, 2026). 7 Giovana Figueiredo Peluso Lopes, Artificial Intelligence and Judicial Decision-Making: Evaluating the Role of AI in Debiasing, 33 TATuP – J. TECH. ASSESSMENT THEORY & PRAC. 28 (2024).
accountability in employing AI. 8 The study conducted by Dr. John mounts importance on ethical governance and judicial education when employing AI in judiciary. It also provides a note on comparative applications of AI between India and China. The work ignores to examine statutory framework or judicial decisions in India by only adhering to the analysis of AI impact on judicial reasoning. 9 This article written by Mr. Siddharth evaluates the significance of judicial accountability and transparency. Yet, it lacks exploration of comparative jurisprudence and ignores the current AI tools in force in Indian judicial system. 10 The work by Mr. Sahibpreet Singh compares AI initiative in India with risk-based regulatory model in EU. The article ignores to fully explore the conflict between AI use efficiency and constitutional principles. Further, the work is policy oriented.
11 Yet another work points to the challenges of algorithmic bias and the want for transparency, citing global examples. However, the research does not critically examine mechanisms for operationalising human oversight, the role of audit trails, or specific ethical frameworks guiding AI-assisted decision-making in Indian courts. It also lacks an integrated discussion of recent judicial policies, such as the Kerala High Court’s AI policy, which is pivotal for practical governance insights. 12
V. AI IN COURTS: GLOBAL FRAMEWORK AND INDIAN RESPONSE
A. Global Perspectives and Emerging Concerns
International scholarship unveils increasing fear about usage of AI in Judiciary. AI in judicial system is categorized by European Union as “high-risk”, requiring safeguards in accountability and transparency.13 Researchers indicate that the algorithmic efficiency cannot surpass fair procedure or jeopardize the reasoning requirement in judicial adjudication.14 It is demonstrated by study that while AI expedites case processing and grants legal research support, advantages 8 Bhaswati Talukdar, An Analytical Study on the Use of Artificial Intelligence in Judicial Decision-Making in India, 6 INT’L J. RSCH. PUBL’N & REVS. 156 (2025).
9 John Varghese, Artificial Intelligence Assisted Judicial Processes—A Primer, SSRN (Oct. 15, 2024),
https://papers. ssrn. com/sol3/papers. cfm? abstract_id=5056102. 10 Siddharth Peter de Souza, AI and the Indian Judiciary: The Need for a Rights-based Approach, The Hindu, Oct. 13, 2024. 11 Sahibpreet Singh, Transcontinental Analysis of Artificial Intelligence-Based Judicial Systems in EU and India, Constitutional Discourse (Nov. 1, 2025), https://constitutionaldiscourse. com/a-transcontinental-analysis-ofartificial-intelligence-based-judicial-systems-in-eu-and-india/. 12 Amal Singh Patel, Algorithmic Bias and the Quest for Equal Justice in India, VIRTUOSITY LEGAL (Sept. 9, 2025), https://virtuositylegal. com/algorithmic-bias-and-the-quest-for-equal-justice-in-india/. 13 Regulation (EU) 2024/1689 of the European Parliament and of the Council of June 13, 2024 Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act), 2024 O. J. (L 1689).
14 Syarifah Lisa Andriati et al, Justice on Trial: How Artificial Intelligence is Reshaping Judicial Decision- Making, 9 J. INDONESIAN LEGAL STUD. 909 (2024).
should be weighed against inviolable constitutional principles. The algorithms cannot be permitted to replace judges without altering fundamentally what “justice” means.15 B. Current AI Initiatives in Indian Courts
The e-courts project phase III undertaken by supreme court of India allocates Rs. 7,210 crore for enhancing the judicial infrastructure across the country with Rs. 53. 57 crore being granted for the purpose of development of AI. 16 SUPACE helps judges by proving summary of case files, factual briefing, and assists in finding the pertinent legal provisions through machine learning. It primarily processes massive information to aid judicial decision making rather than deciding the cases. SUVAS resolves the language barrier by translating major judges of the court, from English language to Hindi, Punjabi, Telugu, Gujarati, and Kannada. This enable the persons not acquainted with English language to understand the decision affecting them. Until 2025, several thousands judgments had been translated into the said regional languages.
Further, LegRAA is a tool used for legal research, ASR-SHRUTI for the purpose of transcribing Constitutional Bench oral submissions into text, and PANINI for the purpose of document translation. 17 C. Promises of Judicial AI
Genuine benefits are offered by responsible AI deployment. The case management with use of AI involves classification of cases on the basis of complexity and age, leading to reduction of delays.18 Machine learning model provide suggestion regarding utility of resource allocation in trial courts and it has demonstrated significant accuracy in the identification of case of delay in lower courts.19 The AI powered search system has made the databases within the reach for normal litigants which otherwise would be accessible only to law-firms owing to the latter’s financial capacity.20 The massive equity gap is addressed by the translation as the rural litigants 15 Olha M. Yarmaki et al., Artificial Intelligence in Judicial Proceedings and Court Decisions: Potential and Risks, 23 VISNYK PRAVO (2023).
16 Nivash Jeevanandam, supra note 2.
17 Use of AI in Supreme Court Case Management, PRESS INFO. BUREAU (Mar. 20, 2025),
https://www.pib.gov.in/PressReleasePage.aspx?PRID=2113224®=3&lang=2. Impact of
AI
on Judicial Decisions in India,
LAWWEB.IN
(Oct. 5, 2025), https://www.lawweb.in/2025/10/judging-machine-impact-of-artificial.html. 19 Mohit Bhatnagar & Shivaraj S. Huchhanavar, Predicting Delays in Indian Lower Courts Using AutoML and Decision Forests, in Computing, Internet of Things and Data Analytics: ICCIDA 2023 166 (Springer 2024). 20 Devendra Singh et al, Application of Machine Learning Based Assumptions in Legal Profession—An Appraisal, IEEE XPLORE (May 8, 2024), https://ieeexplore.ieee.org/document/10593058.
not acquainted with English can understand the judgments affecting them. It benefits the judges to focus solely on substantive legal work by ensuring no additional time is spent on paperwork. D. Documented Risks and Concerns
The conclusion by machine learning algorithm is in the manner indecipherable to humans. This leads to violation of natural justice as the reasoning behind the decision is not explained by the judges in aid of AI. 21 This gives rise to opacity problem. Yet another issue is with bias perpetuation. The research indicates that bail prediction model projects disparities based on influence by religion or caste information existing in the historical data. 22 Reliance on such data institutionalizes the existing discrimination even with the employment of AI for decision making in the judiciary. As many AI systems were made with use of American legal principles, it may not be appropriate for Indian social conditions to adopt the same as it will lead to inaccurate results without consideration of distinct minority rights, depressed classes rights and peculiar approach to family law.
23 The recommendations of AI put the judge under psychological pressure to side with the recommendation of AI without the application of the individual judicial mind. 24
VI. CONSTITUTIONAL AND LEGAL FRAMEWORK
A. Constitutional Provisions
Article 14 states that there is no denial of equality before law or equal protection of laws, to any person by the state. This right is violated when the outcomes are given by the algorithms trained in biased historical data. Such deployment of AI producing discriminative results owing to bias, leads to disparity and amounts to arbitrary action by the state. Article 21 Protects life and personal liberty of persons. It also includes within itself the right to fair trial and speedy justice. In Maneka Gandhi v. Union of India,25 the supreme court held that the procedure must be just, fair and reasonable rather than technically lawful. This
21 Hendrickx, supra note 4.
22 Girhepuje, supra note 5.
23 Lopes, supra note 7.
24 Talukdar, supra note 8.
25 Maneka Gandhi v. Union of India, (1978) 1 S.C.C. 248.
principle is violated by opaque algorithmic influence. In K.S. Puttaswamy v. Union of India,26 the court has propounded three factor test for state affecting privacy: legitimacy, necessity and proportionality. The sensitive personal data is processed by judicial system may lead of violation of privacy rights for the individuals whose data is fed into the AI tool.
B. Statutory Gaps
India’s legal framework for electronic transactions and cybercrime is governed by The Information Technology Act, 2000. This legislation has no provisions for addressing algorithmic transparency or decision making by AI systems.27 It has no specification of provisions for AI system processing sensitive case information such as medical details, family circumstances, financial information disclosed in litigation.28 This gap in India projects the lack of legal framework governing AI deployment in courts. VII. Judicial response in AI adoption
The first judicial reliance on AI was in the case Jaswinder Singh v. State of Punjab29, wherein the court has asked ChatGPT for bail jurisprudence in cruelty cases in a matter relating to a bail petition. The judge stated that the determination had no influence on the decision which was given applying own judicial mind. 30 This indicates the temptation to rely on AI but ensuring the human intervention as being superior to AI. This case indicates that while AI research assistance may be productive to decision making, the human deliberation is essential to comprehend the litigants circumstances for arriving at just decision making. In Christian Louboutin SAS v. The Shoe Boutique31, the Delhi High Court’s reference to ChatGPT demonstrates the limited role of AI in adjudication.
While its responses were considered, the Court clarified that such outputs are merely informational and cannot substitute judicial reasoning. The court reiterated that AI cannot serve as human intelligence or replace humans. 32 While AI can assist with preliminary research, it cannot be engaged for substantive
26 Justice K.S. Puttaswamy v. Union of India, (2017) 10 S.C.C. 1.
27 C. S. Suraj, Regulating Artificial Intelligence in India: Constitutional Challenges, and the Need for a Comprehensive
AI
Law,
INT’L
J.
ADVANCED
RSCH.
SCI.
COMM.
&
TECH.
(2025), https://doi.org/10.48175/IJARSCT-30504. 28 Deepika, Ajit Kumar & Pranjal Sahay, AI and Data Privacy Law: Are We Truly Protected by Our Data Shield?, IISPPR (Feb. 27, 2025), https://iisppr.org.in/ai-and-data-privacy-law-are-we-truly-protected-by-our-data-shield/.
29 Jaswinder Singh v. State of Punjab 2023 PHHC 055530
30 Id.
31 Christian Louboutin SAS v. The Shoe Boutique CS(COMM) 583/2023
32 Id.
judicial determination. In 2025, Kerala High Court has rolled out a formal AI policy that expressly prohibits usage of AI for decision making.33 The usage is permitted only for administrative, research, and translation functions subject to human oversight. This exist as the most detailed AI regulation for the purpose of its usage in subordinate judiciary and furthers institutional goal to preserve human decision making authority as a constitutional imperative.
VIII. CRITICAL ANALYSIS
A. The Assistive vs. Decisional Distinction
The crucial gap exists between AI assistance and AI usage for decision making. The assistance involves summarizing, translation of judgments and alike. The final decisional making is with the human in the case of AI assistance. The risk appears when the AI suggests bail or sentencing quantum, replacing the human mind. The AI usage must be always in assistive category, not in decisional spectrum. The judicial function has to be taken with independent exercise of judicial mind by the human-judge and it cannot be delegated to any algorithm. Such delegation may lead to trifling the constitutional principles of equality, privacy and fair trial.34 B. Bias as Constitutional Problem
Algorithmic bias is distinct from bias in private decision making. The state institutionalizes discrimination through its official justice system when the algorithm discriminates. This directly leads of violation of Article-14. The empirical evidence captures this issue from the studies that found significant disparity in bail prediction model based on AI, indicating the influence of caste or religious identity on outcome.35
C. The Efficiency-Justice Tension
A fundamental tussle exists between the promise of AI in expediting case disposal and the need
for fair judicial process. The efficiency cannot be purchased by sacrificing procedural fairness or human judgment. Constitutional guarantee embody commitment to human dignity which 33 High Court of Kerala, Policy Regarding Use of Artificial Intelligence Tools in District Judiciary (July 19, 2025). 34 Siddharth Peter de Souza, AI and the Indian Judiciary: The Need for a Rights-Based Approach, THE HINDU CENTRE (Nov. 27, 2024).
35 Girhepuje, supra note 5.
cannot be traded for administrative convenience.36
IX. CONCLUSION
AI can be utilized to support judicial administration in assistive capacities such as legal research, translation, case management, and administrative processing. The existing AI tools namely SUPACE and SUVAS project strong utility without replacing human judgment. The constitutional guarantees underlying within Articles 14 and 21 provide foundation to restrict AI's decisional role, but legislative lacunae prevent sufficient implementation. India lacks comprehensive statute to govern the usage of AI in judicial system. The Kerala High court has adopted commendably cautious approach by limiting AI to assistive functions. Kerala's policy and court pronouncements establish that human judgment remains non-delegable.
It is vital to note that the algorithmic bias poses serious constitutional threat that leads to violation of equality and due process in Articles 14 and 21, if the AI systems are deployed without rigorous fairness testing and transparency mechanisms. It is imperative for the parliament to enact comprehensive AI governance legislation establishing transparency standards, fairness requirements, and audit mechanisms for judicial systems. While the efficient judicial system is the need of the hour, the fundamental constitutional values cannot be diminished at any cost. 36 Shreya Tiwari, Towards a Rights-Based AI Framework in India: Bridging Global Models with Constitutional Duties, LIVELAW (July 27, 2025), https://www. livelaw. in/lawschool/articles/towards-rights-based-aiframework-india-bridging-global-models-constitutional-duties-298896 .